Dear most2050,
There are two aspects of the problem; One, to compute the temperature differential across the section and other to check its effect.First issue can be dealt with by heat transfer analysis.If that is difficult , assume a conservative estimate of inside and outside temperatures . Then use SEC III subsection NB where rules govern this case. Whether or not fatigue analysis is required depends on how many such cycles are there.I used this method in a thermal shock analysis in an LNG plant. Of course issues like metallurgical transformations due to rapid cooling rate like martensitic transformation or precipitation hardening are issues which the above solution methods will not address, although they do not appear to be issues significant to your problem.
Regards
Edited by anindya stress (04/04/10 06:43 AM)
_________________________
anindya